Overcoming Barriers to Integrating Behavioral Health and Primary Care Services

Abstract

Objective: Despite barriers, organizations with varying characteristics have achieved full integration of primary care services with providers and services that identify, treat, and manage those with mental health and substance use disorders. What are the key factors and common themes in stories of this success? Methods: A systematic literature review and snowball sampling technique was used to identify organizations. Site visits and key informant interviews were conducted with 6 organizations that had over time integrated behavioral health and primary care services. Case studies of each organization were independently coded to identify traits common to multiple organizations. Results: Common characteristics include prioritized vulnerable populations, extensive community collaboration, team approaches that included the patient and family, diversified funding streams, and data-driven approaches and practices. Conclusions: While significant barriers to integrating behavioral health and primary care services exist, case studies of organizations that have successfully overcome these barriers share certain common factors.

Introduction

Mental disorders are a serious, costly, and neglected public health concern. The World Health Organization (WHO) has identified depression as the leading cause of disability worldwide.¹ In the United States, about half of all adults will have a diagnosable mental illness in their lifetime.² Persons with mental illness have lower life expectancy and higher rates of chronic disease, work absenteeism, unemployment, and poverty.^3-6 Mental disorders are 1 of the top 5 most costly conditions.^7,8 Despite the availability of treatments, only about 45% individuals in the United States with a mental illness received treatment in 2014.⁹

Among individuals with mental illness who do seek care, 56% do so in a primary care setting.¹⁰ Often, primary care providers (PCPs) are inadequately equipped to handle behavioral health issues, including mental illness and substance use disorders, due to lack of training and a fragmented health care system.^11,12 For example, Cunningham¹³ found that two-thirds of surveyed primary care physicians could not get an outpatient mental health service for their patients. Despite this, 65% of all psychotropic medications are prescribed by PCPs.¹⁴ In addition, co-occurring mental and physical illness, or comorbidity, is common, complicating the diagnosis and treatment of both mental and physical illness.^15,16

Integrating primary and behavioral health care can address these gaps. Research has shown improved clinical outcomes for patients receiving integrated care.¹⁷ A systematic review and meta-analysis found collaborative chronic care models of integrated health resulted in significant improvements in depression, mental and physical quality of life, and social role function for patients with a variety of mental illnesses.¹⁸ A 2012 review of collaborative care models, found that these models of team-based primary and behavioral care significantly improved symptoms of depression and anxiety compared with standard primary care.¹⁹ Successful integrative health models within the Veterans Health Administration (VA) and Federally Qualified Health Centers (FQHCs) have shown increased access to behavioral health services, reduced stigma related to receiving behavioral health services, increased patient satisfaction, and improved clinical outcomes.²⁰ In addition to patient outcomes, integrated health models are cost-effective in reducing mental illness symptoms across a variety of populations.²¹

Integrated Health Models

The term integrated health varies across the literature, but most generally refers to the connection of behavioral health and medical services.²² In the most complete models of integrated health, these services are interwoven to produce a tightly integrated, on-site care team who work together to deliver a comprehensive service to their patients. Other models of integrated health include colocation and coordinated care.^23,24 Each model involves a patient-centered approach to care.²⁵

Individual organizations may follow one model of integration or may combine elements of multiple models. For example, a patient-centered medical home (PCMH) is an integrated health model that includes coordinated, team-based care for both primary and behavioral health needs. The PCMH often builds care management and behavioral health consultation into the model.²⁶ Another example is the delivery of behavioral health services through primary care, that most commonly includes behavioral health screenings and brief interventions (SBI) and referral to community resources by PCPs. SBI includes the recognition of alcohol and substance use via structured guidelines, such as flowcharts,²⁷ and the subsequent direction of patients to management or specialty care.²⁸ In addition, behavioral health providers can assist PCPs through collaborative care in which providers, case managers, and consulting psychiatrists, combine perspectives to address each patient’s needs and remain in communication to continue care plan revision or modification while under care. Collaborative care models can range from PCPs consulting by telephone, interactive media, or mobile devices with outside mental health specialists, education of PCPs by behavioral health clinicians, and structured care based on disease management principles.²⁹

Barriers to Implementation

Despite evidence supporting integrated health models, there has been minimal implementation of comprehensive models across the United States. Translating integrated models to a clinical setting from research is challenging due to reimbursement issues, limited capacity, resistance to change, information technology issues, and confidentiality rules for behavioral health.³⁰ Previous work by Grazier et al³¹ also found that several key barriers hindered wider adoption. Their systematic literature search identified a variety of barriers, falling into several key categories: a focus on vulnerable populations (for example, mental health is a secondary concern when treating children with HIV), patient and family factors (for example, lack of culturally competent mental health care providers to treat ethnic minorities), comorbidities (for example, providing adequate care for individuals with multiple physical comorbidities and mental illness), provider factors (for example, perceived doubts of their ability to implement integration), financing and costs (for example, lack of reimbursement for care management services), and organizational issues (for example, provider shortages).³¹ This article sheds light on the possible means to improve the pace of integrated health implementation across the United States by investigating programs that have had success in overcoming known barriers to implementation.

Organizational Achievement

This article recognizes that organizational “achievement” or “success” as used in the traditional academic and business literature relies primarily on profitability and high financial returns.³² However, health services research on organizations in the public and nonprofit sectors have influenced thinking and research on the importance of organizational culture in the sustained success of organizations, whether investor owned or community-owned. As Cameron³² notes, “The sustained success of these firms has had less to do with market forces than company values, less to do with competitive positioning than personal beliefs; less to do with resource advantages than vision.” There is a comprehensive literature of systematic reviews and original research on what causes organizational success and failure; thorough examination and important works related to varying theories and methods of inquiry include industrial organization approaches and industrial ecology perspectives.^33-37 Whether sustained organization performance, one measure of success, is due to internal or external factors has over two decades of exploration. Most recently, health services delivery research has identified the importance of strong and pervasive leadership, a champion for the “cause,” the importance of networks of support, and of course, financial feasibility as crucial factors in sustained performance and basic tenets of organizations in the midst of changing communities and policies.³⁸ While an in depth analysis of the organizational behavior literature addressing success and failure, in various business sectors, is well beyond the scope of this article, we adapt several consensus determinants from these literatures in the conceptual framework for our analyses.

This study sought to identify characteristics of organizations that have successfully integrated mental health and primary care. We posited that similar themes would emerge across even dissimilar organizational models.

Methods

Depending on the rubric, the methods for this study are often categorized first, as social science research methods, second, as evaluation research, and third, as qualitative methods. Among these frameworks, this study is considered field research, using observation and one-on-one interviewing. The study has some characteristics of mixed methods, particularly when accompanied by the review of literature and organization and community-specific financial and legal information.³⁹ As Creswell⁴⁰ notes, however, research practices lie somewhere on a continuum between quantitative and qualitative.

Based on a comprehensive literature review of integrated models of care for mental health and substance use, and a nonprobability based snowball (or referral) sampling technique to identify additional practice sites, the study identified 30 organizations that had been able, for at least 3 years, to integrate care.⁴¹ (The literature review was the product of extensive database searches conducted with the assistance of the reference librarians (informaticians) from the WHO Library, National Library of Medicine, and the University of Michigan libraries: Taubman Medical; School of Public Health; School of Social Work; College of Engineering. The review and summary of the peer reviewed and agency-generated (e.g., SAMHSA, NIMH) literature used key search terms to derive definitions of “integrated” and “integration;” organizational and systems change; the breadth of potentially integrated services (for example housing, juvenile justice, criminal justice, schools, mental health and addiction treatment systems, primary and specialty care); services integration; integrated treatment; co-occurring illness integrated treatment; addiction; evidence-based practices for behavioral health treatment.) After conducting telephone interviews with directors, presidents, chief executives, or senior program managers of each of these 30 organizations, 6 were scheduled for 3 to 5 day visits, due to their replicability, comprehensiveness, and representation of differing models of integration.

Sampling and Site Selection

One of the goals of this approach was to identify initiatives, programs, or systems that demonstrate an integrated approach to solving multiple, complex, and serious substance abuse, mental health, medical, and social problems faced by individuals and families. We imposed selection criteria for sites, requiring that they serve at a minimum the vulnerable family populations targeted in this study, and that they illustrate the variation in the organization, locus of control, governance, financing structure, and other key dimensions of interest. Sites could link specialized delivery systems through formal networks, provider partnerships, interagency agreements, and related means to create a locus of accountability and a services delivery system for these vulnerable families. Others could have coordinated but separated and/or categorically tied financing streams, and develop purchasing standards or formulae for reimbursement or risk sharing. These entities illustrate differences in auspices, goals, organizational structure, age, size, economic leverage, and consumer focus. What binds them together is the goal to deliver integrated care to families who have serious coexisting health and social problems that impact morbidity, mortality, and quality of life.

Data Collection

The purpose of the interviews was to generate the data needed to describe decision-making/policy-making processes in each site comprehensively, in detail, and from multiple perspectives. While the interviews had a clear purpose, they were “minimally structured,” with a list of topic areas and some specific questions to be asked of most of the respondents. They were “nondirective” to some extent, to allow the control of the conversation to be shared by interviewer and interviewee.

More than 300 individuals were interviewed in person or by telephone from these 6 sites, in addition to almost 100 thought leaders; managers; researchers; community advocacy group leaders; consumers; participants and leaders of treatment support groups; chief financial, executive and operating officers of small and large, rural and urban organizations; foundation program officers; and treatment providers. The project gathered quantitative and qualitative data on organizational histories, structures, financing, management processes, legal requirements, enabling authorities, contracts, and other critical topics. The question asked of almost all respondents was that related to the “active” or “essential” ingredients for initiating, designing, implementing, evaluating, and sustaining integrated services for vulnerable populations. There were no refusals to be interviewed.

Two authors jointly developed a set of inductive codes after initial review of the data, and then independently coded themes from on-site notes from each site visit. The 2 authors discussed and mutually agreed on the addition of emergent codes and resolution of all discrepancies.

This study was approved by the University of Michigan’s Institutional Review Board. (IRB #HUM00012872).

Results

Our analysis of case studies of organizations that successfully integrated behavioral health into primary care identified 6 broad factors shared by most or all organizations that may have helped them overcome previously identified barriers. The 6 shared practices were a focus on vulnerable populations; use of data-driven practices; community-wide collaboration; presence of a person or persons with deep institutional vision; a team approach to care that included the patient and family or caregivers; and diverse funding streams. The 6 organizations included in this analysis varied in organization structure, scope, and location (Table 1).

Table 1.

Case Study Characteristics.

Organization	Location	Organization Type	Scope of Integrationa
Behavioral Health and Recovery Services	San Mateo, California	Public health entity of San Mateo County	Integration of public mental health and substance abuse services to better treat co-occurring disorders
Community Caring Collaborative	Washington County, Maine	Coalition of state agencies and local organizations	Integrated care to better treat infants and children in households affected by substance abuse
Community Partnership of Southern Arizona	Southern Arizona	Nonprofit agency; State of Arizona Regional Behavioral Health Authority	Integration of behavioral health, medical care, and social services for enrolled members
Denver Health	Denver, Colorado	Public health entity of the City and County of Denver	Integration of behavioral health and medical care for the safety net population
Intermountain Healthcare	Salt Lake City, Utah	Nonprofit health system	Delivery of mental health services in the primary care setting
New Mexico Behavioral Health Collaborative	Santa Fe, New Mexico	State initiative	Delivery of behavioral health services in federally qualified health centers

^a“Scope of Integration” includes a description of integration practices derived from the on-site notes from each site visit.

Prioritizing Underserved Vulnerable Populations

Successful organizations often targeted “vulnerable” populations identified as underserved or in need of additional services. For example, Behavioral Health and Recovery Services (BHRS) of San Mateo County prioritized delivery of services to homeless families and individuals; New Mexico’s Interagency Behavioral Health Purchasing Collaborative’s Veterans and Family Support Services initiative targeted military personnel, veterans, and their families,⁴² and the Community Partnership of Southern Arizona deliberately targeted those women who were pregnant or postpartum and those with HIV/AIDS. Many models, such as BHRS⁴³ and Denver Health also focused on incarcerated populations, with Community Partnership of Southern Arizona staff embedding staff into the Pima County juvenile justice system to provide services to detained or incarcerated youths.

Use of Data-Driven Best Practices

Reliance on data-driven best practices to guide organizational strategy was a common theme among organizations. Intermountain used clinical, patient satisfaction, and cost outcomes data to evaluate, test, and revise the Mental Health Integration (MHI) program and conducted data-driven needs assessments to determine where additional community partnerships were needed. Denver Health was an early adopter of health information technologies^44-46 and conducted periodic internal audits to assess whether racial or ethnic treatment disparities existed in its system.⁴⁷ Many organizations also prioritized implementation of established best practices, including collaboration with the community and utilization of a team approach that included the patient, detailed below.

Community-Wide Collaboration

Another theme to emerge was community-wide collaboration. The forms of collaboration were unique to each organization, but each relied heavily on support from other entities as well as the community as a whole. New Mexico’s Collaborative included representatives of 21 state departments or offices, ranging from the Office of Workforce Training and Development to the Governor’s Commission on Disabilities,⁴⁸ and prioritized local needs by creating 13 local collaboratives to represent districts of the state as well as 5 additional collaboratives to represent the state’s Native American population.⁴⁹ Maine’s Community Caring Collaborative consisted of 37 member organizations, including medical centers and hospitals, early education programs, colleges and universities, and a variety of nonprofits and local and state agencies. The Community Partnership of Southern Arizona collaborated with the community in part through its board of directors, which included service recipients and their families, and received community input from a Public Policy Committee composed of service recipients, family members, providers, and other members of the community. In addition to including patient members, BHRS of San Mateo County’s Steering Committee also included members from partner organizations or county departments working on related issues.

Support From Influential Leaders and Established Institutions

Several organizations were championed by influential supporters and received strong support from established institutions. For example, the creation of the BHRS was driven by a collection of high-level leaders from the San Mateo County Health System, the San Mateo County Human Services Agency, and the California Mental Health Service Division. The Collaborative was authorized in 2004 by the Governor, with state legislature support, as a response to complaints about the lack of quality and high cost of the current state public behavioral health services.

Team Approach That Includes the Patient and Family

All models included in this analysis utilized a team-based approach and included the patient and the patient’s family as a part of the care team. For example, Intermountain’s MHI teams were comprehensive and holistic, including the patient and his or her family, as well as a care manager, physician, psychologist, other mental health specialist, and clinic administrator. Similarly, the Community Partnership of Southern Arizona created a team to guide each patient’s care. This team consisted of the patient, those he or she identified as a support network, and a clinical liaison.⁵⁰ San Mateo operated under a “Welcoming Framework,” a key piece of which is patient involvement in addressing problems.

Diverse Funding Streams

Each organization had a unique funding structure, but all were diversified to varying degrees. Many received partial funding through Medicaid, with Denver Health receiving a substantial share of its funding through disproportionate share hospital payments and several other organizations receiving funds directly from state Medicaid programs. State general fund dollars and grant funding were also key components of several programs’ funding structures. In-kind funding was an essential element of several organizations’ funding strategies. Maine’s Community Caring Collaborative relied heavily on community support in the form of volunteered space and personnel time as well as other in-kind donations, and Intermountain’s partnerships with local government and schools yielded donations of clinic space.⁵¹

As an integrated system including a health plan, Intermountain was uniquely positioned to evaluate the effects of integrated behavioral health and primary care services on the overall cost of care. Rather than requiring additional expenditures, MHI instead cut Intermountain’s costs through reductions in emergency department utilization, psychiatric admissions, and inpatient length of stay.⁵²

Barriers and Keys to Success

The themes that emerged from this analysis matched well with the barriers previously described by Grazier et al.³¹ Each identified theme corresponded to one or more barrier, indicating the potential to serve as a facilitator to organizations seeking to overcome a specific barrier. Barriers and the corresponding factors leveraged by the organizations studies are presented in Table 2.

Table 2.

Barriers and Accommodations for Barriers.

Barrier	Accommodations
Vulnerable populations	• Prioritization of vulnerable populations
Patient and family factors	• Community-wide collaboration • Team approach that includes the patient
Comorbidities	• Prioritization of vulnerable populations
Provider factors	• Team approach that includes the patient
Financing and costs	• Diverse funding streams
Organizational issues	• Use of data-driven best practices • Presence of a “champion”

Discussion

Individuals with mental illness often have poor health outcomes, incur high costs, and suffer due to inadequate care.^{3,4,8,9,11,13} Integrated health care models have repeatedly shown success in improving care.^17-20,23 Our findings suggest that successful organizations share certain commonalities that may have contributed to their ability to overcome or accommodate those barriers, including prioritization of underserved vulnerable populations, use of data-driven best practices, community-wide collaboration, support from influential leaders and established institutions, reliance on a team approach including patients and family members, and diversification of funding streams. These results add to the literature regarding barriers that create a gap between research and clinical implementation of integrated health models, illuminating potential target areas that may encourage the integration of primary care and behavioral health. Further work is needed to determine whether these commonalities are unique to organizations that have successfully integrated behavioral health and primary care services or whether they are shared by less successful organizations. Despite the potential lack of generalizability of case studies, our results suggest that health organizations can and do succeed at integrating behavioral health and primary care services. Organizations considering integrating behavioral health and primary care services may use the current findings as a guide to achieve these goals.

Conclusion

While significant barriers exist in integrating health models of behavioral and primary care, the six organizations included in this analysis successfully integrated behavioral health and primary care. Commonalities among these organizations included prioritizing underserved vulnerable populations, increasing community collaboration, ensuring strong leadership early in the process, implementing a team-based approach including the patient as an active participant, diversifying funding sources, and implementing data-driven best practices.

Author Biographies

Kyle L. Grazier is Richard Carl Jelinek Endowed Professor of Health Services Management and Policy, Chair of the Department of Health Management and Policy at the School of Public Health, and Professor of Psychiatry at the Medical School at the University of Michigan.

Mary L. Smiley, MS, is a Health Policy Analytics Fellow at the Center for Healthcare Research & Transformation. She conducts research on behavioral health and US health insurance.

Kirsten S. Bondalapati, MPH, is a Health Policy Fellow at the Center for Healthcare Research and Transformation. Her work includes research and policy analysis related to healthcare reform and mental health.